Bug 68575 - [hsw/bdw] *ERROR* Timed out waiting for forcewake old ack to clear.
Summary: [hsw/bdw] *ERROR* Timed out waiting for forcewake old ack to clear.
Status: CLOSED WORKSFORME
Alias: None
Product: DRI
Classification: Unclassified
Component: DRM/Intel (show other bugs)
Version: XOrg git
Hardware: Other All
: medium normal
Assignee: Ben Widawsky
QA Contact: Intel GFX Bugs mailing list
URL:
Whiteboard:
Keywords:
: 72773 (view as bug list)
Depends on:
Blocks:
 
Reported: 2013-08-26 14:16 UTC by Chris Wilson
Modified: 2017-07-24 22:57 UTC (History)
3 users (show)

See Also:
i915 platform:
i915 features:

Attachments

Description Chris Wilson 2013-08-26 14:16:26 UTC 
There is a low probabilty of this timeout ocurring, especially noticeable if polling the rings for GPU busyness.
Comment 1 Daniel Vetter 2013-08-27 08:43:20 UTC 
(In reply to comment #0)
> There is a low probabilty of this timeout ocurring, especially noticeable if
> polling the rings for GPU busyness.

So essentially it's broken, only pampered over by us not using forcewake too often? Do consistency checks of the forcewake register turn up anything? Iirc I've heard some noise that this register sometimes drops writes or wreaks other havoc ...
Comment 2 Paulo Zanoni 2013-11-01 15:46:31 UTC 
Is this something we can try to reproduce with an IGT test?
Comment 3 Paulo Zanoni 2013-12-04 20:37:02 UTC 
"drm/i915: use the correct force_wake function at the PC8 code" solves the bug I could reproduce. The way to reproduce it was through our sysfs i915_forcewake_user file, so this is probably not the bug Chris was mentioning...
Comment 4 Ben Widawsky 2014-01-10 20:53:15 UTC 
Chris, close?
Comment 5 Chris Wilson 2014-01-11 12:09:49 UTC 
If you want to file a HSD... :-p
Comment 6 Ben Widawsky 2014-01-11 20:34:41 UTC 
Do we see the problem for extraordinarily high values of FORCEWAKE_ACK_TIMEOUT_MS?
Comment 7 Ben Widawsky 2014-01-14 00:01:28 UTC 
*** Bug 72773 has been marked as a duplicate of this bug. ***
Comment 8 Ben Widawsky 2014-01-23 18:01:41 UTC 
Chris, do we see this with extremely high FORCEWAKE_ACK_TIMEOUT_MS too?
Comment 9 Chris Wilson 2014-01-24 12:05:37 UTC 
Just tried to reproduce this on crw without much luck. I might try again on the original hsw that showed the issue later on.
Comment 10 Ben Widawsky 2014-01-28 01:10:37 UTC 
I vote let's close it for now, and you can reopen when you see it?
Comment 11 Ben Widawsky 2014-01-29 19:13:26 UTC 
Jani, would you be able to come up with a list of potential commits?
Comment 12 Ben Widawsky 2014-01-29 19:16:26 UTC 
Crap, modified wrong bug. Sorry.
Comment 13 Ben Widawsky 2014-03-25 03:51:29 UTC 
Chris does this still occur? With the fixes from Mika, it appears to be gone on GEN8 at least.
Comment 14 Chris Wilson 2014-03-25 16:51:29 UTC 
No, I can no longer easily trigger this.

Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.